Η Google πρόσθεσε κάποιες νέες λειτουργίες ασφάλειας στο Gmail, και μια από αυτές φέρεται να βελτιώνει την ανίχνευση ηλεκτρονικού “ψαρέματος” ή phishing χρησιμοποιώντας σε ένα εξειδικευμένο μοντέλο μηχανικής μάθησης.
According to the company, 50 to 70 percent of incoming messages to the Gmail are spam, but machine learning techniques help Google to beat spam and phishing emails with an accuracy of 99.9%.
The latter machine learning model improves the process by delaying selected messages (less than 0.05% of messages on average) for further fishing analysis.
Google says that the new detection models will also help to create new warnings delaying URL access time by showing users a warning question when they have clicked on a suspicious link. As new patterns are identified, new protection models will be adapted and improved over time.
In addition, Google added external reply warnings to the company's email to prevent users from inadvertently sending protected data to someone outside their company.
Gmail will not stop sending but will try to delay it a bit with warnings, according to Google, as machine intelligence will notify Gmail if the recipient is an existing contact or someone the sender interacts with regularly.
Google has also added new built-in defenses against ransomware and polymorphic malware, blocking millions of add-ons that circulate attached to emails.
Threats are identified by combining flags from spam, malware and ransomware with euristics attachments (emails that may be flagged as threats based on the flags) and the digital signatures of the senders (already flagged as malware).